Two Halves are More than One: Phase-wise Velocity Distillation for Fast and High-Quality Image Generation

Zhen Guo; Rongyuan Wu; Qiaosi Yi; Chenxi Xie; Xinyu Wei; Lei Zhang

cs.CVOct 6, 2026

Two Halves are More than One: Phase-wise Velocity Distillation for Fast and High-Quality Image Generation

Authors: Zhen Guo, Rongyuan Wu, Qiaosi Yi, Chenxi Xie, Xinyu Wei, Lei Zhang

Organizations: Department of Computing, The Hong Kong Polytechnic University · OPPO Research Institute

Abstract

Recent diffusion-based image generation backbones have grown substantially in scale, making the network inference cost increase rapidly. While diffusion distillation techniques can reduce the number of inference steps, high-quality image generation within a single full-backbone-forward compute budget remains challenging. Existing one-step methods typically allocate this budget to a single evaluation of a monolithic student. However, approximating the heterogeneous coarse-to-fine transport with a single monolithic mapping is difficult and often leads to over-smoothed outputs. To address this issue, we propose Phase-wise Velocity Distillation (PVD), which partitions the generation timeline into a coarse and a fine phase, and models the transition within each phase via the average velocity. A dedicated half-sized expert is assigned to each phase, decoupling structural composition from detail refinement while keeping the cumulative computation equivalent to one full-backbone forward pass. We show that the use of two half-sized phase-specific experts outperforms a single full-size monolithic student. On class-conditional image generation, PVD achieves an FID of 1.48 on ImageNet 256 x 256. On more complex text-to-image (T2I) tasks, PVD-distilled models (Stable Diffusion 3.5-Medium, FLUX.1-dev, Qwen-Image) produce results competitive with their multi-step teachers, significantly outperforming prior distillation methods. Moreover, across the evaluated T2I backbones, PVD reduces active parameters by 49.10-50.89% and peak VRAM by 45.76-48.36% compared to the corresponding teachers. Source code and distilled models are available at https://github.com/PolyU-VCLab/PVD.

Figures & tables

Figure 1 : Text-to-image results. Top : Teacher model (Stable Diffusion3.5-Medium, FLUX.1 dev and Qwen-Image under $50\times 2$ , $50$ , and $50\times 2$ steps, respectively). Bottom : With two half-sized phase-wise experts, PVD achieves fast and high-quality image generation using cumulative computation equivalent to one full-backbone forward pass of the teacher model.

Figure 2 : Trajectory-based distillation (left) and distribution-based distillation (middle) employ a uniform student model to approximate the full coarse-to-fine mapping, while our proposed phase-wise velocity distillation (right) decomposes the objective into phase-wise velocity targets and sequentially evaluates two half-sized experts under a single full-backbone-forward compute budget.

Figure 3 : Overview of Phase-wise Velocity Distillation (PVD). (a) A half-sized backbone is extracted from the teacher and fine-tuned via flow matching to initialize the local students. (b) Phase-specific LoRA experts are distilled by mean velocity learning across partitioned time intervals. (c) For T2I tasks, phase-wise discriminators are used to provide adversarial refinement.

Table 4

Model	GenEval	DPG	WISE	QI	TIIF-Bench		Aesthetic	Pick	Image	$N_{\mathrm{flops}}$	Active Params.	Peak VRAM
Model	GenEval	DPG	WISE	Bench	short	long	Score	Score	Reward	$N_{\mathrm{flops}}$	Active Params.	Peak VRAM
SD3.5 Medium [ 4 ]	0.6829	84.49	0.45	43.19	0.7166	0.7190	5.7212	21.45	0.6127	100.00	2.24	4.99
LADD [ 26 ]	0.0941	26.46	0.05	6.55	0.2232	0.2019	4.3533	18.40	-2.0724	1.00	2.27	4.86
DMD2 [ 28 ]	0.4563	46.15	0.31	29.27	0.5410	0.4931	4.7885	20.09	-0.1003	1.00	2.24	4.81
TwinFlow [ 58 ]	0.6171	72.81	0.37	30.55	0.5856	0.6094	5.0111	20.26	-0.0154	1.00	2.25	4.83
SenseFlow [ 30 ]	0.5222	81.98	0.31	29.43	0.6273	0.5957	4.9707	20.69	0.4834	1.00	2.24	4.81

Table 3 : Quality and efficiency of distilled text-to-image generation. The right two columns report active parameters (B) and peak VRAM (GiB), together with PVD’s reductions relative to its teacher model. Higher is better for quality metrics; the best and second-best results among distilled methods within each backbone group are highlighted in bold and underlined , respectively.

Figure 4 : Visual comparisons of T2I generation. Compared with existing distillation methods, PVD preserves much clearer structure and finer details under a single-forward budget, demonstrating close T2I generation quality to its multi-step teachers.

Figure 5 : Visual comparisons after further training on Unsplash data. The models further trained on Unsplash images are marked with ∗ , which can generate more authentic photorealistic images with improved lighting, shadows, composition, and fine-grained details.

Backbone	Weights	Aesthetic score $\uparrow$	DreamSim diversity $\uparrow$	GenEval $\uparrow$
SD3.5-Medium	PVD	6.4753	0.1473	0.6948
SD3.5-Medium	PVD ∗	6.5993	0.1832	0.6697
FLUX.1-dev	PVD	6.4471	0.1457	0.6532
FLUX.1-dev	PVD ∗	6.7941	0.2238	0.6334
Qwen-Image	PVD	6.6463	0.1079	0.8846
Qwen-Image	PVD ∗	6.7289	0.1432	0.8594

Table 4 : Results of further training on Unsplash data. The models further trained on Unsplash images are marked with ∗ . Better results per backbone are highlighted in bold .

Appendix figures & tables19 assets

Supplementary material from the paper’s appendix.

Appendix

	Transport factorization				Inference efficiency
Method	Phase-restricted training	Phase-specific parameters	Average-velocity objective	Phase-factorized inference	Reduced active params. & VRAM	Few-step distillation
MeanFlow	$\times$	$\times$	✓	$\times$	$\times$	✓
Hierarchical Distill.	$\times$	$\times$	Init.	$\times$	$\times$	✓
TimeStep Master	✓	✓	$\times$	$\times$	$\times$	$\times$
Glance	✓	✓	$\times$	$\times$	$\times$	$\times$
Phased DMD	✓	✓	$\times$	✓	$\times$	✓

Appendix

Table 5 : Comparison with representative related methods in terms of temporal localization of the training objective, parameter specialization, average-velocity modeling, and explicit phase-factorized inference. “Init.” indicates that average-velocity learning is used only for student initialization rather than retained as the final phase-specific transport objective.

Table 6 : Training configurations for PVD. Backbone initialization and phase-expert training are reported separately. Weight decay is

0

and gradient clipping is

1.0

Figure 6 : More class-conditional samples generated by PVD.

Model	Single Object	Two Object	Counting	Colors	Position	Attribute Binding	Overall $\uparrow$
SD3.5 Medium	0.9875	0.8131	0.6469	0.8324	0.2400	0.5775	0.6829
LADD	0.2812	0.0025	0.0625	0.2181	0.0000	0.0000	0.0941
DMD2	0.9000	0.3636	0.4750	0.6489	0.1200	0.2300	0.4563
TwinFlow	0.9594	0.6490	0.5688	0.7527	0.4250	0.3475	0.6171
SenseFlow	0.9344	0.6212	0.4562	0.6888	0.1625	0.2700	0.5222
SWD	0.9656	0.7601	0.5844	0.7793	0.2625	0.4200	0.6286

Appendix

Table 7 : Quantitative comparison on GenEval. Higher is better (

\uparrow

) for all metrics. The best and second-best accelerated results within each backbone group are in bold and underlined , respectively.

Model	Global	Entity			Attribute					Relation		Other		Overall
Model	Global	Whole	State	Part	Color	Shape	Size	Tex.	Oth.	Spat.	N-Spat.	Cnt.	Txt.	Overall
SD3.5 Medium	91.04	90.44	90.07	88.36	90.36	89.01	89.99	90.17	89.28	91.15	89.46	91.03	87.40	84.49
LADD	35.63	40.89	49.58	60.31	58.88	51.69	38.16	47.68	47.23	54.96	57.50	60.65	51.86	26.46
DMD2	62.69	49.92	61.15	42.04	54.09	32.37	54.49	54.74	56.69	39.39	57.86	60.65	47.09	46.15
TwinFlow	82.16	83.64	82.10	83.57	86.04	89.33	82.89	80.87	82.32	85.47	83.95	84.87	81.22	72.81
SenseFlow	88.56	87.29	87.86	85.11	89.30	91.55	90.30	84.80	86.92	89.23	89.12	87.65	88.66	81.98

Appendix

Table 8 : Quantitative evaluation on DPG-Bench. Higher is better (

\uparrow

) for all metrics. The best and second-best accelerated results within each backbone group are in bold and underlined , respectively.

Model	Cultural	Time	Space	Biology	Physics	Chemistry	Overall
SD3.5 Medium	0.43	0.50	0.52	0.41	0.53	0.33	0.45
LADD	0.04	0.03	0.07	0.04	0.05	0.07	0.05
DMD2	0.28	0.28	0.40	0.25	0.44	0.30	0.31
TwinFlow	0.27	0.40	0.53	0.40	0.50	0.38	0.37
SenseFlow	0.27	0.32	0.43	0.33	0.39	0.22	0.31
SWD	0.45	0.27	0.45	0.30	0.35	0.10	0.36

Appendix

Table 9 : Quantitative comparison on WISE. Higher is better (

\uparrow

) for all metrics. The best and second-best accelerated results within each backbone group are in bold and underlined , respectively.

Model

TIIF Overall

Spatial (2D)

Spatial (3D)

Action

Color

Texture

\uparrow

\uparrow

(L/S)

Col

Tex

Col

(L/S)

SD3.5 Medium

71.66

71.90

83/83

79/70

91/86

82/82

73/65

68/75

95/90

70/50

92/88

90/85

83/77

92/100

LADD

22.32

20.19

0/41

29/41

8/13

17/12

26/34

20/27

40/40

35/20

16/8

28/19

33/50

36/56

DMD2

54.10

49.31

75/ 87

62 / 79

72/80

76/ 87

61/ 73

72 / 68

95 / 95

70/70

52/72

76/ 100

72/66

80/80

TwinFlow

58.56

60.94

79/ 91

79 /75

75 / 88

84 / 84

57/ 73

68/ 75

90 / 95

86 / 75

72/84

85 / 95

77 / 77

84/88

Appendix

Table 10 : Quantitative comparison on TIIF Benchmark mini (Part I: Visual & Relational). L/S stands for Long/Short. Scores are multiplied by 100. Higher is better (

\uparrow

) for all metrics. The best and second-best accelerated results within each backbone group are in bold and underlined , respectively.

Model

Comparison

Differentiation

Negation

Base

Col

Tex

Base

Col

Tex

Base

Col

Tex

(L/S)

SD3.5 Medium

79/75

68/75

62/41

66/66

30/20

84/76

56/62

82/70

80/61

75/85

70/62

68/81

64/64

61/72

72/66

LADD

16/29

0/0

6/6

4/38

15/30

52/48

0/0

0/5

47/52

10/25

50/50

56/50

35/47

72/44

44/33

DMD2

66/ 70

75 /50

25/62

61/ 90

33/20

88 /80

43/50

84 /64

71/61

25/ 70

66 /54

56/ 75

64/ 76

72/ 77

61 / 55

TwinFlow

79/ 83

50/ 75

31/ 68

90 /57

70 / 40

92 / 84

87 / 93

70/70

71/ 90

55/ 75

62 /58

75 / 75

58/58

77 /61

55 / 55

Appendix

Table 11 : Quantitative comparison on TIIF Benchmark mini (Part II: Logic & Reasoning). L/S stands for Long/Short. Scores are multiplied by 100. Higher is better (

\uparrow

) for all metrics. The best and second-best accelerated results within each backbone group are in bold and underlined , respectively.

Model

Numeracy

Shape

Real World

Style

Text

Base

Col

Tex

Col

Tex

(L/S)

SD3.5 Medium

72/70

77/48

78/87

65/77

70/59

45/40

65/70

78/86

72/68

75/73

70/66

48/67

LADD

22/12

7/22

25/40

22/30

31/9

25/15

20/20

26/36

24/16

27/23

3/3

0/0

DMD2

50/64

55/ 70

46/71

47/60

38/54

45/30

55/60

66/ 88

58/66

46/42

16/13

0/5

TwinFlow

68/ 70

62/55

65/71

60 /57

47/56

60 / 55

80 / 80

72/ 90

64/ 78

57/55

33/26

3/6

Appendix

Table 12 : Quantitative comparison on TIIF Benchmark mini (Part III: Concepts & Domains). L/S stands for Long/Short. Scores are multiplied by 100. Higher is better (

\uparrow

) for all metrics.The best and second-best accelerated results within each backbone group are in bold and underlined , respectively.

Model	Quality	Aesthetics	Alignment	Real-world	Creative	Overall
Model	Quality	Aesthetics	Alignment	Fidelity	Generation	Overall
SD3.5-Medium	47.77	45.88	41.83	42.08	33.05	43.19
LADD	1.22	7.19	4.41	30.44	0.54	6.55
DMD2	22.68	35.84	30.18	36.90	19.89	29.27
TwinFlow	29.21	37.13	28.83	36.03	18.54	30.55
SenseFlow	30.26	33.04	27.60	36.35	17.83	29.43

Appendix

Table 13 : Quantitative comparison on Qwen-Image-Bench. Higher is better (

\uparrow

) for all metrics. Best results within each distillation group are in bold , and the second-best results are underlined .

Model	Compute		Model runtime		Model footprint		Change vs. teacher
Model	$N_{\mathrm{flops}}$	TFLOPs	Latency (s)	Throughput (images / s)	Active / Total Params. (B)	Peak VRAM (GiB)	Active Params.	Peak VRAM
SD3.5 Medium	100.00	701.93	15.49	0.07	2.24 / 2.24	4.99	Ref.	Ref.
LADD	1.00	7.02	0.16	6.69	2.27 / 2.27	4.86	$\uparrow$ 1.34%	$\downarrow$ 2.61%
DMD2	1.00	7.02	0.15	6.89	2.24 / 2.24	4.81	0.00%	$\downarrow$ 3.61%
TwinFlow	1.00	7.02	0.16	6.67	2.25 / 2.25	4.83	$\uparrow$ 0.45%	$\downarrow$ 3.21%
SenseFlow	1.00	7.02	0.16	6.67	2.24 / 2.24	4.81	0.00%	$\downarrow$ 3.61%

Appendix

Table 14 : Detailed hardware-cost comparison. The final two columns report changes relative to the corresponding teacher;

\downarrow

and

\uparrow

denote reductions and increases, respectively.

Figure 7 : Qualitative comparisons on SD3.5-Medium, FLUX.1-dev, and Qwen-Image. Rows are grouped by backbone and labeled with the corresponding teacher or distillation method.

Figure 8 : Additional visualizations after further training on Unsplash data. The models further trained on Unsplash images are marked with ∗ .

Panel of Figure 9 — Figure 9 : Coarse-to-fine progression of PVD distilled from Qwen-Image. Within each pair, the left image visualizes the state after the first phase and the right image shows the output after the second phase. The first phase reveals coarse composition, approximate subject silhouettes, and color of regions despite residual noise. The second phase refines this structure with sharper edges, surface textures, and finer letter strokes.

Method	$N_{\mathrm{flops}}$	Active Params. (B)	Total Params. (B)	TFLOPs	Peak VRAM (GiB)	GenEval $\uparrow$
Qwen-Image	100.00	20.43	20.43	5585.14	38.42	0.87
MeanFlow	4.00	20.44	20.44	223.40	38.48	0.44
MeanFlow	8.00	20.44	20.44	446.80	38.48	0.49
PVD (Ours)	1.02	10.40	10.55	56.81	19.84	0.88

Appendix

Table 15 : Reported quality–compute comparison on Qwen-Image. GenEval results for MeanFlow are sourced from [ 58 ] .

Configuration	$N_{\mathrm{flops}}$	Active Params. (B)	Total Params. (B)	TFLOPs	Peak VRAM (GiB)	GenEval $\uparrow$
w/o discriminator	0.99	1.10	2.22	6.93	2.67	0.5599
Full-size student	1.00	2.24	2.24	7.02	4.99	0.4142
Full-size student	2.00	2.24	2.24	14.04	4.99	0.5695
PVD	0.99	1.10	2.22	6.93	2.67	0.6948

Appendix

Table 16 : Component and compute-control ablations on SD3.5-Medium.

Phase 1 $\rightarrow$ Phase 2	GenEval $\uparrow$	ImageReward $\uparrow$
$E_{\mathrm{early}}\!\rightarrow E_{\mathrm{late}}$	0.6948	0.4947
$E_{\mathrm{early}}\!\rightarrow E_{\mathrm{early}}$	0.6754	0.4598
$E_{\mathrm{late}}\!\rightarrow E_{\mathrm{late}}$	0.1820	-1.7190
$E_{\mathrm{late}}\!\rightarrow E_{\mathrm{early}}$	0.2346	-1.5811

Appendix

Table 17 : Expert specialization and execution-order ablations on SD3.5-Medium. Rows specify the expert used in the first and second trajectory phases, respectively.

Panel of Figure 10 — Figure 10 : Prompt: A top-down studio photograph of exactly thirty-five small pink macarons arranged in a precise rectangular grid of five rows and seven columns on a dark slate tray. Every row contains exactly seven macarons and every column exactly five. All macarons are identical, separate, fully visible, and not touching. The entire tray is inside the frame. Both methods violate the requested count and grid dimensions despite producing recognizable, regularly arranged objects.

Panel of Figure 11 — Figure 11 : Prompt: A clear studio photograph of one upright white ceramic mug with its handle on the right. A long red pencil passes diagonally through the empty hole of the mug handle, from upper left to lower right. The pencil is outside the cup, never entering the opening at the top. Both ends of the pencil extend visibly beyond the handle. Three-quarter front view showing the handle hole and the empty cup opening, on a plain gray tabletop. Both methods place the pencil through the cup opening and mug wall instead of the existing handle hole.

Explore similar work

Parallel Decoding Distillation for Fast Image and Video Generation
Jul 28, 2026Neta Shaul, Chao Liu, Arash Vahdat +1Few-Step Distillation Dataset Distillation
2Weizmann Institute of Science · 1NVIDIA
High-Fidelity Two-Step Image Generation via Teacher-Aligned End-to-End Distillation
Jun 10, 2026Dongyang Liu, Ruoyi Du, David Liu +7Multi-Reference Image Generation Few-Step Generation
Z-Image Team, Alibaba Group
One-Step Distillation of Discrete Diffusion Image Generators via Fixed-Point Iteration
May 20, 2026Chaoyang Wang, Yunhai TongDiffusion Models Probe-Logit Distillation
Peking University

Jul 28, 2026, cs.CVJ/K move · Enter open · S save

Parallel Decoding Distillation for Fast Image and Video Generation

Neta Shaul, Chao Liu, Arash Vahdat, Julius Berner

2Weizmann Institute of Science · 1NVIDIA

Few-Step Distillation Dataset Distillation

Open paper AskPDF arXiv

Generation in video diffusion or flow models is computationally expensive due to the slow and iterative sampling process. Current state-of-the-art (SOTA) acceleration methods heavily rely on variational score distillation (VSD) and adversarial losses to distill diffusion models into few-step generators. Albeit achieving high-quality video generation, these training losses are notoriously hard to optimize and suffer from mode collapse, leading to loss of video diversity and lack of motion. In this paper, we introduce Parallel Decoding Distillation (PDD), a simplified and scalable trajectory-based distillation method for fast inference of diffusion and flow matching models. Our architecture and training procedure are compatible with any pre-trained model and support sampling with a varying number of function evaluations (NFE). PDD accelerates generation by predicting multiple denoising steps per network evaluation. Conceptually, it learns a representation of the mean velocity without regressing its derivative using JVPs or finite-difference approximations. Our method achieves SOTA performance with 4-8 NFE on LTX-2.3 Text-to-Video/Audio, Wan 14B Text-to-Video, and Qwen-Image Text-to-Image. Moreover, PDD presents a significant improvement in generated video diversity.

Abstract

Explore similar work

Parallel Decoding Distillation for Fast Image and Video Generation

High-Fidelity Two-Step Image Generation via Teacher-Aligned End-to-End Distillation

One-Step Distillation of Discrete Diffusion Image Generators via Fixed-Point Iteration